HPE Primera

2020: Basic autonomy from the person in ensuring availability

On June 16, 2020 the Hewlett Packard Enterprise (HPE) company announced updates of the storage systems (SS) of HPE Primera and HPE Nimble Storage within development of the Intellectual platform for data. It is a solution portfolio, relying on the artificial intelligence (AI), developed for work in a multicloud environment and provided to customers on model "as service". In the provided updating to opportunities of DWH it is added: for Primera — basic autonomy from the person in ensuring availability; for Nimble — support of drives of data of Storage Class Memory (SCM, memory of a class of storage).

According to the developer, the updated opportunities include:

• self-optimization of system processes thanks to AI which is built in HPE Primera. The storage system is ready to possible failures of different components, in advance building optimal ways of isolation of problems and continuation of normal operation of applications;
• expanded cross to the analyst in HPE InfoSight for the platform virtualizations Microsoft Hyper-V. As well as earlier for VMware, InfoSight can visualize the arising delays and to make recommendations about performance improvement at all levels infrastructures, from components of DWH Nimble to separate virtual machines;
• disaster recovery after failure thanks to a possibility of replication on three DPCs in HPE Primera and HPE Nimble and also fast asynchronous replication for HPE Primera;
• support of the HPE Primera configurations with disks only the NVMe interface or the mixed configurations of NVMe-SAS which increase performance of applications on certain profiles of loading;
• support of the modules SCM in HPE Nimble which are capable to provide decrease in response time twice for a number of architecturally compatible applications;
• support of virtual volumes of VMware vVols for Primera for automation of selection of capacity under requests from virtual machines (Nimble support vVol).

As noted in HPE to accelerate development and to increase stability of the business, the companies look for an opportunity to quickly bring new applications to the market, continuing to support already existing at the same time. However IT departments hardly cope with new tasks as they are busy with support of crucial applications and data. As a result, they are forced to sacrifice flexibility for the sake of stability of work. HPE Primera and HPE Nimble help not to make this choice, unloading administrators and providing permanent availability of applications. And possibilities of this DWH for flexible selection of resources and their payment in process of use with the HPE Greenlake program, give the chance quickly and with minimum risks to implement new services and applications.

According to the statement of the developer, HPE InfoSight, AI for support of infrastructure, helps IT specialists to focus on business development, predicting and preventing failures both in hardware components, and in virtual environment. HPE allocated HPE Primera with an opportunity to work independently, without the need for online connection to InfoSight to ensure continuous and fast functioning of crucial applications. The architecture of HPE Primera includes the AI built-in engine trained at global models from "big" InfoSight that in real time to predict behavior of applications and pattern of performance. On the basis of the local analysis of telemetry of Primera dynamically optimizes selection of resources for ensuring the predicted performance.

One of the main opportunities of HPE InfoSight — the analyst of the events at all levels of infrastructure, from "iron" to virtual machines, is available also to virtualization of Microsoft Hyper-V (for June, 2020 only for DWH of Nimble). The cross analytics of telemetry from different parts of an IT range reveals anomalies in productivities and problems arising between DWH and virtual machines more precisely and also non-optimal used resources, consider in HPE.

For ensuring protection of applications against catastrophic events of scale of the whole IT platform or even larger, it is necessary that data of applications were stored in several geographical distributed places. Primera includes HPE Peer Persistence updated software which provides a possibility of synchronous replication of data to two reserve platforms at once. For example, it can be reserve DPC in the adjoining building and the rented platform for several tens of kilometers (the remoteness is defined by delays in the channel). Such two-level protection, according to the developer, allows not only to be protected from the large-scale failure influencing at once two platforms but also to provide the different service levels in case of disaster recovery for different applications. For example, crucial data can be vosstanalivat closer to users, and minor to take out on cheaper remote platform. In addition to it, in Primera asynchronous replication on long distances with target recovery point (RPO) 1 minute became available.

According to the statement of HPE, purchasing HPE Nimble, the companies receive the platform with availability at the level of "six nine" on one platform, HPE Peer Persistence with automatic recovery after failure on two other platforms (as well as in Primera) and a possibility of asynchronous replication on long distances on own DWH or in a cloud. Both on HPE Primera, and on HPE Nimble, replication can be configured quickly without special skills of work with DWH, additional expenditure for software and need for professional service. All updates are included in the basic license "all inclusive" according to the policy of HPE, for their activation it is enough to update a firmware of DWH. In HPE Primera the multinodal architecture which provides a low delay at 75% of transactions of input-output at the level of 0.25 ms of HPE Primera is put supports the All-NVMe configuration that allows is more complete to use its mass and parallel architecture for performance improvement. For example, for SAP HANA HPE Primera in the All-NVMe configuration supports twice more nodes.

According to HPE, SCM realizes high-speed performance potential at the level of RAM, but with complete non-volatility. However, the drives SCM cost much for permanent so far storages of all operational data. HPE Nimble, thanks to the architecture of CASL developed for caching of an internal file system showed good read performance in hybrid SSD/HDD configurations. In June, 2020 in HPE announced optimization of architecture of CASL for caching on SCM to accelerate the hybrid SCM/SSD configurations. It allows to lower average delay twice to level no more than 0.25 ms, with the minimum influence on the price of DWH.

As noted in HPE, the companies from all industries transform methods of development and deployment of applications, applying automation at all stages. Virtual volumes (vVol) optimize transactions of resource management of storage using the services focused on dynamic virtual machines. After HPE Nimble, HPE Primera also supports vVol and provides a full range of opportunities of vVol, including disaster recovery after failure with support of Site Recovery Manager.

For containers with the applications saving the status (stateful) the companies need storage, supporting software-defined automation. HPE Primera, as well as HPE Nimble, supports the CSI driver for Kubernetes to allocate developers with opportunities for dynamic management of data storage. Container platforms, such as HPE Container Platform, thus become integrated with HPE Primera at the sufficient level for crucial corporate applications.

The release of the HPE Primera OS 4.2 operating system will be available worldwide in the 3rd quarter 2020 free of charge to customers of HPE Primera having current contracts on support.

As of June, 2020 are already available to the order worldwide:

• HPE Primera NVMe;
• The HPE CSI driver for Kubernetes 1.1.1 for HPE Primera;
• Sets of SCM drives of 1.5 of Tb for HPE Nimble for existing or new a flash arrays of HPE Nimble AF60 or AF80;
• Replication on three DPCs for HPE Nimble is available as a part of release of NimbleOS 5.2 to any customer of HPE Nimble having the current contract on support;

The cross analytics of HPE InfoSight for Hyper-V will be available worldwide in July, 2020 to any customer of HPE Nimble having the current contract on support.^[1]

2019

Integration with HPE Synergy

On November 4, 2019 the Hewlett Packard Enterprise (HPE) company announced integration of crucial elements of the portfolio which combine possibilities of the artificial intelligence (AI) and the arranged infrastructure. The company integrated the platform for HPE Primera data storage relying on AI with the arranged HPE Synergy platform to help customers to unroll quicker new applications and to develop innovations for development of the business.

This combination, according to the developer, will allow customers to unroll own infrastructure with accurate SLA on availability of services, permanent expansion of functionality and management via flexible programmable interfaces in the spirit of "infrastructure-as-code".

As noted in HPE, the modern companies aim to accelerate everything: work of the applications, removal on the market of new products and services and reorganization of infrastructure under the current tasks. Regardless of whether the application works at purely equipment rooms virtualized or containerized resources, IT specialists need to provide fast and continuous functioning of infrastructure. A variety of tasks of the modern enterprise dictates need to manage any types of resources as to a whole by the uniform principles and tools.

"Customers want something bigger, than technologies for basic program-controlled automation of virtualization. HPE proposes the arranged solution combining program-controlled automation of any types of resources and the intellectual platform of storage of any data types. It allows to make work in local infrastructure more similar to work in a cloud", 'Mat Mathews, the vice president and the CEO on the arranged HPE networks noted'

The arranged infrastructure

According to the statement of HPE, the arranged infrastructure of HPE opens opportunities for scaling of resources for calculations, storages and data transmissions independently of each other, but at the same time provides the uniform mechanism of management with them and their distributions between any loadings and the service levels in the mode real time as required that allows to reduce as the capital investments, and to save huge arrays of time of administrators which can be redirected on creation of innovations. The uniform application programming interface (API) HPE Synergy allows developers to integrate directly management of "iron", virtualization and containers into development processes, testings and deployments of applications and also into third-party applications and tools.

Intellectual infrastructure

According to the developer, the intellectual platform of data storage HPE Primera uses possibilities of HPE InfoSight, the large-scale knowledge base about functioning of different environments. This layer of information including telemetered informations from tens of thousands the customer worldwide is automatically analyzed using artificial intelligence. Such analysis allows to fix in real time the arising problems arising at some customers and to prevent them at all others and also to extend the best practices on all similar environments. Use of InfoSight shows good results: with it customers of HPE have an opportunity to reduce costs of time for systems management of data storage for 93% and to increase performance of applications. HPE Primera whose architecture has deep integration with InfoSight opportunities, gives to customers the chance to create innovations, without going at the same time on compromises, due to fault tolerance and efficiency of storage in combination with flexible use of cloud computing which are necessary for business.

According to the developer, inclusion of HPE Primera in a solution portfolio for the arranged infrastructure provides to the customers using HPE Synergy, following features:

• Flexibility and fast time-to-market
  • the Possibility of more rapid development and deployment of applications using approach "infrastructure-as-code" which allows to automate and optimize resource management. For example, reduction of time for deployment databases Oracle reaches 89% (according to internal tests of HPE).
  • Elimination of interruptions and dead time on liquidation of problems thanks to work of AI on all IT infrastructure.
  • Self-sufficiency and an opportunity to eliminate delays at accomplishment of requests.

• High fault tolerance
  • Thanks to AI, HPE InfoSight, available in servers HPE ProLiant, HPE Synergy and storage systems of HPE Primera, predicts and prevents problems before their emergence.
  • Reliability during the work with the crucial systems is put in HPE Primera. The platform provides high availability thanks to multinodal architecture and technology of replication of data HPE Peer Persistence. Besides, Primera becomes covered by the "Guarantee of 100% Data Availability" program.
  • HPE provides additional protection against interruptions of service and risks using the technology of check and recovery of microcodes of HPE Silicon Root of Trust which prevents loading of a system with a komprometirovanny component of the server and also thanks to recognition within the Cyber Catalyst program which allows to receive more advantageous conditions of insurance upon cyberthreats at the insurance companies participating in the program.

• Smooth and dynamic scaling
  • the Solutions HPE for the arranged infrastructure allow the companies to arrange as necessary resources independent from each other without work interruption.
  • Possibilities of smooth automatic detection and autotune promote forward scaling and integration of resources.
  • Quick access to data on use of resources applications gives the chance to better plan loading.

Thanks to a combination of opportunities of AI and the arranged infrastructure, HPE creates for customers the intellectual platform for work in "cloud" style even on the own platform, claim in HPE.

Release of the HPE Primera platform

On June 19, 2019 the Hewlett Packard Enterprise (HPE) company announced HPE Primera release — platforms which offer essentially new opportunities of data storage in the crucial systems and provide ease of use, availability and performance. HPE Primera relies on HPE InfoSight opportunities - the platform of artificial intelligence providing the automated support of the equipment. It allows HPE Primera to show good results: the platform reduces costs of time for systems management of data storage for 93%, allows to predict and prevent emergence of problems and increases performance of applications, claim in HPE.

For support of business development, increase in flexibility and acceleration of the innovation processes the modern companies should bring new crucial applications to the market and support already existing. However IT departments, as a rule, do not manage to conform to these requirements as they are overloaded with administration tasks, adjustments and supports of infrastructure. As a result they are forced to sacrifice flexibility IT infrastuktury for its stability.

According to the statement of the developer, the HPE Primera platform combines innovations of all product portfolio and solutions for data storage of HPE and reinterprets possibilities of crucial storage systems.

According to the developer, the HPE Primera platform does data storage in the crucial systems to the simplest and available:

• Simplicity of installation. The customer can set independently HPE Primera less than in 20 minutes.
• Simplicity of management. To select capacity to business applications of data to the HPE Primera platform, seconds are required. Storage optimization function (compression and a compression) which if necessary can be disconnected is built in the platform.
• Simplicity of updating. The built-in operating system allows to initialize, update and restart services of data management separately. The customer can update the software of HPE Primera in only five minutes, without interrupting system operation.
• Simplicity of operation. Customers can use the HPE Primera platform on model "as service", paying only that capacity which is required, through HPE GreenLake. The customers purchasing a product in a full ownership receive the Timeless Storage tool excluding need of complete replacement of the equipment. The platform allows to perform replacement of the controller without stopping works and offers the full-function software, compression and data availability.

According to the developer, HPE Primera guarantees 100% availability of crucial DWH:

• Guarantee of 100% availability. HPE Primera includes a service package of HPE Proactive Care and a guarantee of 100% data availability. It is not necessary to sign special contracts or to incur additional costs.
• Control of borders of data warehouses. More than 90% of problems arise out of data warehouses — at the level of networks, servers and virtualization. The HPE Primera platform allows to control operability of infrastructure outside data storage level thanks to the HPE InfoSight system which predicts and prevents failures at the level of network functioning, applications and servers.
• Stability for the crucial systems is built in the platform. HPE Primera provides high availability at the level of architecture thanks to support to four controllers, HPE Peer Persistence technology, is transparent supporting business work, and replications of data. The platform includes the system of data protection of the HPE Recovery Manager Central applications which in combination with HPE StoreOnce increases data recovery speed by 15 times and provides economical long-term data storage in a public cloud by means of HPE Cloud Bank Storage technology.

The expected performance for unforeseen loadings:

• High performance at consolidation of applications with high system requirements. The architecture of all-active of an array of HPE Primera provides a parallelization of input-output operations that increases performance of applications. Work of Oracle accelerates for 122%, claim in HPE.
• Safe consolidation of applications. The HPE InfoSight technology and the artificial intelligence which is built in an array provides automatic adjustment of an array for optimization of work of a set of the business applications storing the data on HPE Primera.
• The analytics in real time allows to support the expected performance. The HPE InfoSight technology is even more deeply integrated into HPE Primera. The mathematical models based on artificial intelligence and trained in a cloud are built in an array and provide analytics of real time for stabilization of loadings of applications.

"The intelligence changes everything. The modern enterprises need absolutely new approach to maintenance of infrastructure of operating activities of business — based on the intellectual strategy of work with data. HPE Primera relies on the HPE Intelligent Data Platform platform and provides ease of use, availability and high performance of applications for the crucial systems. It will allow the companies to focus on use of data for development of their business", noted Milan Shetti, the senior vice president and the CEO of HPE on storage systems

The HPE Primera platform is included into the product portfolio and the solutions HPE Intelligent Data Platform intended for performance improvement of applications, improvement of data management, flexible use of cloud computing and expansion of opportunities of business at the expense of analytics of real-time data. Intelligent Data Platform offers customers not only storage, but also the intellectual data processing allowing to reveal the hidden business value. HPE Intelligent Data Platform includes the platform of infrastructure management on the basis of artificial intelligence and machine learning of HPE InfoSight which forms an intellectual basis of HPE Primera and all other products HPE for data storage which represent the full-scale system of offline storage of data based on artificial intelligence.

The HPE Primera platform will be available to the order in August, 2019.

Notes